Abstract: On rapid data of large-scale network abnormal database localization in the process of research, using the current algorithm when positioning rapid data cannot be accurately extract the specific characteristics of abnormal data exist the problem of low positioning accuracy data. Is proposed based on improved fertility group of firefly algorithm of large fast positioning data modeling method of network anomaly database. To form a large network anomaly database entity model, and on the basis of the model to form a detailed analysis of the causes of the formation of abnormal data, extracting network anomaly characteristics of the abnormal data in the database at the same time, the fertility group of firefly algorithm, to obtain large network anomaly characteristics of the specific abnormal data in the database, and according to its characteristics to establish a large database of power network fault positioning rapid data model. Experimental simulation show that the firefly algorithm based on improved fertility group of large-scale network abnormal database orientation of rapid data modeling method of high precision, strong adaptability.